ABSTRACT As average life expectancy has increased, so has the global prevalence of age-related dementias; Alzheimer's Disease (AD) is a devastating disease that will affect 16 million Americans by 2050. AD is both highly heritable and highly polygenic, with as many as 30 susceptibility loci identified through genome-wide association studies (GWAS), meta-analyses of GWAS, exome array based rare variant analysis, and several early whole exome sequencing (WES) studies to date. Still, much of the genetic architecture of AD remains unknown. Because interpretation of associated variants is difficult devoid of a biological context, the functional role of most associated variants has yet to be described. Re-analysis of existing data resources using functionally oriented analysis approaches represent a novel strategy that will aggregate evidence of risk around gene action to provide new insights into the functional role of associated genetic variants and identify novel genetic mechanisms of risk. Therefore we propose to perform extensive reanalysis of existing genetic, transcriptomic, and imaging data in the following specific aims. 1) Enrich discovery and functional characterization of risk- associated variants in AD using sample diversity and existing tissue-specific expression prediction models. We will estimate gene-based tissue-specific expression in ~90,000 AD cases and controls (~33% cases and ~66% controls) from 62 studies using existing models to identify genetically regulated expression (GReX) associated with AD risk (Aim 1a). We will utilize ancestrally diverse existing GWAS data to refine understanding of the functional variation contributing to gene-level associations (Aim 1b). 2) Generate and make publicly available novel AD-specific models of relevant tissue expression and neuroimaging outcomes for genomic analyses. Using multiple publicly available datasets, we will construct additional AD tissue-specific expression and multi-ethnic reference panels, and make these models publicly available (Aim 2a). We will also implement the GReX methodology within the HAIL high-throughput genomics framework, and explore potential improvements to the approach (Aim 2b). Using available neuroimaging variables, we will also apply the GReX methodology toward the estimation of the genetic component of brain structural features (Aim 2c). 3) Evaluate clinical significance in AD-associated genes. We will develop and apply approaches to identify the broader health outcomes associated with dysregulation of AD-associated genes based on published literature (Aim 3a) and discoveries made in Aims 1 and 2 (Aim 3b) using large DNA databanks such as BioVU (N=100,000). We will examine clinical impact of AD-associated loci phenome-wide by linking genetic variants to the full medical history captured in health systems. Together, these complementary approaches will lead to identification and validation of genes and pathways contributing to risk of AD, providing significant insight into a very common, highly heritable, and debilitating disorder to develop new knowledge and propose expedited strategies for translating genomics to function and clinical utility.